OpenAI unveils Sora: Transforming text into dynamic videos

Currently in the red teaming phase to detect potential flaws, Sora is also accessible to visual artists, designers, and filmmakers for feedback purposes, according to the company's statement.

"Sora excels at crafting complex scenes involving multiple characters, distinct motions, and precise details of subjects and backgrounds. It can even produce various shots within a single video," the company elaborated. Additionally, Sora can bring still images to life, as detailed in a company blog post.

This video generation tool is the latest innovation following OpenAI's ChatGPT, a chatbot launched in late 2022 that garnered attention for its ability to draft emails, write code, and compose poetry. Meanwhile, social media behemoth Meta Platforms has enhanced its image generation model, Emu, with new AI capabilities for editing and creating videos from text.

Facing competition from tech giants like Microsoft, Google, and Amazon, Meta aims to carve out its niche in the fast-evolving generative AI sector. Although still under development, Sora occasionally struggles with accurately interpreting spatial details from prompts and maintaining a specified camera path.

OpenAI is also working on tools to identify videos generated by Sora, underscoring its commitment to innovation and the responsible use of AI technology.